Animals, Mental Defectives,
and the Social Contract
Tom Huffman
Teikyo Westmar University
LeMars, Iowa
and defective bumans or that contractarianism is true,
but one may not consistently believe both.
I wish to sbow two things. First, one may not
rationally resolve this problem one way for defective
bumans and another way for animals. Regardless of
one's substantive position on this dispute, wbatever
reasons one offers regarding one group will apply
equally, ceteriusparibus, to the other. Second, whatever
methodological inclinations one bolds toward intuition­
ism, one may consistently accept a contractarian account
of morality and believe that we bave direct duties to
tbe parties in question. I will explain bow this
compatiblist position is possible, indeed quite plausible.
Wbatever its merit as an account of political
institutions, the social contract theory bas, in recent
years, been the subject of intense scrutiny as an
explanation for our moral obligations. Prominent
examples of this approacb include the work of Jobn
Rawls and David Gauthier. l
A criticism of this understanding of moral obligation
arises from several commentators, partly as a result of
claims made by the original authors. The criticism is
this: The social contract theory seems to imply that the
scope of our moral obligations is unreasonably narrow.
In particular, at fIrSt glance the theory seems to exclude
from direct consideration the interests of animals and
severely defective bumans.
Criticisms of this sort are intuitive in nature and
common to pbilosophy, yet many view them with
disdain. It is an open question whether the failure of an
otherwise attractive theoretical account to square with
our moral intuitions tells against thetheory or the
intuition, but resolving this dispute lies beyond the scope
of this paper. However, surely the intuition and the
theory cannot both be true, or at least this is bow it has
seemed to the majority of commentators on the question.
One either believes that we bave direct duties to animals
Between the Species
1. The Problem
Tbe reason that contractarian accounts of morality
appear inconsistent with possessing direct duties to
beings at the margin of the moral community is fairly
clear. If we understand moral obligations to be the set
of rules that rational, self-interested beings would
accept under a bypothetical set of circumstances, then
sucb rules will be construed by and for rational
creatures. Ordinarily, we enter into contracts for the
purpose of bringing benefit, directly or indirectly, only
to the contractors. If some beings are excluded in
principle from participating in an agreement, and if
the contractors lack a fundamental interest in their
PHILOSOPHY
20
Winter 1993
Animals, Mental Defectives, and the Social Contract
welfare, then it is unclear how the terms of an
agreement could be understood as directly benefitting
them (although they may reap unintended, collateral
benefit). Rawls writes:
explain why it is immoral to exploit animals strictly for
our own purposes and possibly to their detriment.
Last of all, we should recall here the limits
of a theory of justice. Not only are many
aspects of morality left aside, but no account
is given of right conduct in regard to animals
and the rest of nature....While I have not
maintained that the capacity for a sense of
justice is necessary in order to be owed duties
of justice, it does seem that we are not
required to give strict justice anyway to
creatures lacking this capacity.2
The most obvious solution to this dilemma is simply
to grasp one of its horns. I do not intend to offer any
sustained defense or refutation of contractarian theories
of moral obligation, nor would the scope of this paper
allow it. Let us assume, therefore,for the sake of the
argument, that some version of contractarianism is true,
thereby eliminating one solution to the problem.
This leaves, of course, the other option, where we
deny that we have any direct moral obligations to
animals and severely defective humans. Doing this,
however, runs counter, at least in the latter instance,
to some of our most deeply held moral convictions. If
we accept that we have no direct duties to the severely
retarded. then, as Peter Singer points out, we are
committed to "holding that mental defectives do not
have a right to life and therefore might be killed for
food ... or for the purpose of scientific experimen­
tation."6 This is indeed an unsavory position and as
such ought to be avoided as long as rational
alternatives exist.
Intuitive appeals of this sort have lost favor among
philosophers in recent years, but fortunately we need
not rely solely on intuition to address the point. There
are sound theoretical reasons for claiming that severely
defective humans, at least, are appropriate objects of
direct moral concern. The simple fact that they have
awareness of their surroundings, some sense of
continuity of self through time, and an ability to
experience pleasure and pain are all reasons sufficient
to explain their membership in the moral community.
They are subjects of biographical lives, and because of
this we have moral duties toward them.
Another possible solution lies in rescuing our
intuitions by separating animals from severely defective
humans. If there is a significant moral difference
between them, then we could accept contractarianism
and avoid the counter-intuitive results imagined above.
Of course, this option requires at least two things. First,
it must be explained how human yet nonrational
creatures are included within the social contract.
Second, it must be explained how the ftrst explanation
applies to severely defective humans but not to at least
some animals. In other words, what possible moral
difference exists between them?
2. Three Possible Solutions
Interpreting this passage is difficult insofar as
Rawls goes on to say that it is nonetheless "wrong" to
be cruel to animals because of "duties of compassion."
Rawls seems to say that his is only a theory of justice
and not a comprehensive theory of moral obligation
and that we may have direct moral obligations to
animals that go unexplained by his theory. However,
insofar as the concept of nonmaleficence is contained
within the concept of justice (it is incoherent to
suppose that one could be both just and cruel), 3 the
theory would be awkward at best if it presumed to
explain the latter and not the former. 4 Nonetheless,
Rawls clearly sees tension between contractarianism
and one's presumed duty of nonmaleficence to non­
rational creatures.
In a similar vein, David Gauthier writes:
Only beings whose physical and mental
capacities are either roughly equal or mutually
complementary can expect to fmd cooperation
beneficial to all. Humans benefit from their
interaction with horses, but they do not
cooperate with horses and may not benefit
them .... We may condemn all coercive
relationships, but only within the context of
mutual benefit can our condemnation appeal
to a rationally grounded morality.s
Aside from the odd comment that we "may not"­
cannot, might not, ordinarily do not?-benefit horses
(by giving a horse open pasture, sugar cubes, and
protection, do we not benefit it?), Gauthier makes it
clear that a "rationally grounded morality" cannot
Winter 1993
21
Between the Species
Animals, Mental Defectives, and the Social Contract
Neither of these tasks is easy. I shall postpone a
discussion of the first approach, but it is fairly clear
that even if we can account for direct moral obligations
to defective humans under contractarianism, we cannot
do so in such a way that excludes nonhuman animals.
What possible morally relevant difference exists
between them? In the case of fetuses or newborns, we
might appeal to their potential rationality, but this
response, although excluding animals, does not address
the severely retarded whose potentialities are quite
limited. The only other candidates are an appeal to
religious doctrine or species membership. Although the
former allows for a real moral distinction, it is doubtfully
true? The latter is certainly true, but allows for no moral
distinction. Several authors have attempted to explain
why species membership simpliciter is morally
significant, but none of the attempts succeed. 8 They all
employ one of two tactics. Either one sees species
membership as morally significant because it represents
a family relationship that persons share with the less
fortunate members of the species, or some species are
deemed morally significant because their members
typically share certain morally important properties. On
either understanding, we owe other humans more moral .
consideration than we do members of other species
Let John and Mary be two individuals that are
not rational beings, and that belong to different
species, but which are indistinguishable with
respect to their psychological capacities and
their mental lives. It is possible that John
belongs to a species 99% of which are rational
beings, while Mary belongs to a species of
which only 1% are rational beings. If species
membership is morally relevant. .. , it will be
wrong to kill John, but may very well not be
wrong to kill Mary.... 9
Tooley's far-fetched thought experiment reveals a
purely theoretical objection to the theory of moral
obligation in question. Any attempt to define the moral
community according, not to the capacities that
individual prospective members possess, but, rather,
according to the groups to which they belong, is
unsatisfactory. If we start down the road of including
beings within our scope of moral concern simply
because they are members of some group or another,
we may just as easily exclude others because of their
group membership. This strategy is, of course, typical
ofracism, sexism and all of the other 'isms' which seek
to assign individuals less moral value than others based
on their membership in some group.
One might object, noting that species membership
is merely a sufficient, rather than a necessary, condition
for inclusion within the moral community; hence, we
do not risk this type of exclusion. However, this
response misses the point. The issue concerns how mere
species membership serves any normative role in our
moral reasoning, and adopting one explanation over
another fails to account for how species membership is
relevant in the frrst place.
There is, of course, at least one sense in which an
appeal to group membership is appropriate in this
context. There is nothing at all objectionable in
recognizing that we have moral duties to all and only
those beings who are members of the group defined as
"those possessing morally relevant properties." But this .
understanding, although unobjectionable, completely
begs the question. Group membership may always be
appealed to in the trivial sense that those possessing
morally relevant properties can be recognized as
constituting a group. The difficulty lies in determining
which properties are morally relevant, and I know of
no remotely plausible explanation for the moral
relevance of species membership.tO
regardless ofany other propenies each may possess.
Understanding moral obligation in terms of family
relationships pOses a difficulty for any attempt to specify
just which family groupings are significant. The
defenders of this approach have species in mind, but
one could repeat the same reasoning and insist that we
ought to treat members of our own race better than
members of other races due to the family resemblance
of skin color. This type of argument could be repeated
for a potentially infmite number ofclassifications. Note
that a defender may not argue that some classifications
have moral relevance and others do not because,
according to this approach, we assign moral worth only
after we have decided upon a classification.
The second approach recognizes that the possession
of various morally relevant properties (sentience,
rationality, etc.) is what explains moral worth, but in an
oblique way. This perspective recognizes the moral
value of beings who belong to any classification whose
members typically possess the necessary properties.
This theory of moral worth avoids the problem
encountered by the first attempt, but it runs afoul of an
entirely different problem. Michael Tooley illustrates
the difficulty:
Between the Species
22
Winter 1993
Animals, Mental Defectives, and the Social Contract
apply. However, contractarian morality tries to provide
a general explanation of moral obligation that does not
depend on the presence of any specific prior affections.
Contractarian morality asks us to assume that, while
signing the hypothetical contract, one has no particular
friends or family to consider. We have only rational self­
interest, and from this perspective it is radically unclear
why one would closely identify one's own interest with
the interests of nonrational creatures and thus have any
moral obligations toward them.
On the other hand, the altruistic account offers a
ready explanation for including the nonrational within
the scope of moral concern. We care for animals and
defective humans simply out of a spirit of compassion.
This account provides the needed explanation to be sure,
but it is one clearly inconsistent with contractarianism.
Contractarian morality is intentionally heartless. It
attempts to explain moral obligation within a purely
self-interested, nonaltruistic framework.
The parental model thus fails. Either it fails to
provide the required moral explanation, or it does so at
the expense of the very theory it purports to exemplify.
It seems, therefore, that if we wish to account for moral
obligations owed to the nonrational by the rational
within a contractarian framework, we must find an
explanation elsewhere.
The upshot is that if we wish to maintain that
defective humans have moral value and deserve our
protection, then we are rationally bound to take the same
position relative to those nonhuman animals whose
cognitive capacities are at a similar or greater level of
development. All that remains is a strictly empirical
matter of determining which animals possess the
necessary capacities, a task better suited to cognitive
psychology than moral philosophy.
3. A Promising Approach
One model suggests a possible solution. It is
common for one party to enter into a contract with
another for the purpose of serving, not one's own
welfare, but the welfare of a third party, such as when
parents establish a contract with a day care center for
the purpose of serving a child's interest. This type of
situation is quite common and suggests a promising
approach to the current problem. Perhaps we ought to
think of animals and severely defective humans as
appropriate objects of direct moral duty much as we
think of children in this way. Young children lack the
capacity to make agreements, and yet rational beings
regularly assume contractual obligations toward them
at the request of other rational beings with an
understandable interest in their own children's welfare.
In the example above, the day care operator enters
into the contract solely for the purpose of serving her
own interests. The parents' reasons are perhaps, although
easy to understand, somewhat difficult to explain.
Two explanations seem plausible. Either the
parents identify their own interests with that of the
child's so intimately that their motive is, in an indirect
fashion, also based in self-interest, or natural parental
affection motivates a spirit of altruism. According to
the former explanation, parents are self-interested. The
latter explanation suggests that they are self­
sacrificing. Whichever explanation we accept, the
parental model fails to explain how, according to
contractarianism, we have direct moral obligations to
nonrational beings.
Let's consider each account in turn. According to
the self-interest model, we must imagine that rational
beings would identify with mental defectives and
animals. Why would this be true? Of course, in those
cases in which an ordinary adult human has a special
relationship with either-as in the case of a pet or a
severely retarded child-the self-interest model might
Winter 1993
4. A Solution
Having ruled out a rejection of contractarianism, an
attempt to deny all direct moral duties to animals and
defective humans, an attempt to affirm moral duties to
defective humans while denying the same to nonhuman
animals, and a parental model of contractarian morality,
the only remaining alternative lies in a compatiblist
position explaining how contractarianism might account
for an inclusion of both within the moral community.
As stated at the outset, most commentators on the
subject have rejected this approach, preferring instead
to emphasize incompatibility. To appreciate the strength
of this position, we should review the reasons that
motivate it. Contractarianism seems to exclude non­
rational creatures from the moral community because,
according to the theory, the moral community is defined
in terms of those who participate in the contract. This
is understandable. According to the political model,
individuals come together out of self-interest and live
under the terms of a contract. Society includes only
those who explicitly or implicitly agree to the terms of
23
Between the Species
Animals, Mental Defectives, and the Social Contract
the contract. Similarly, on the moral model, individuals
agree to live under a set of moral rules comprising the
entirety of the rights and obligations to which persons
are subject. The moral community is comprised of only
those who, in a hypothetical circumstance to be
specified, agree to its terms. Since only rational beings
can make agreements, only rational beings are subject
to and protected by the rules that result.
If a contractarian account of moral obligation
includes animals and defective humans, then it must
not follow too closely in the above respect its political
analogue. There are, in any case, good reasons not to
understand a moral theory strictly in terms of a political
model. A political, social contract necessarily includes
rather specific facts concerning human nature and the
natural surroundings. The simple fact that the globe
includes scores ofindependent political systems testifies
to the importance of ethnic, geographical, and historical
facts in binding us together in a bewildering number of
ways. According to the moral model, though, such
contingencies are presumably less important. Indeed.
Rawls' account in particular asks us to imagine a
hypothetical original position in which such facts are
not known. The moral rules that result exist on a much
higher level of abstraction than corresponding political
rules of the traditional social contract.
Appreciating this relatively high level of abstraction
helps explain how the interests of nonrational beings
can be included within a Rawlsian-type contract. Those
in the original position are rational and self-interested,
but there is no guarantee that their status remains so
after emergence from behind the veil of ignorance. The
capacities necessary for rational choice possessed by
those behind the veil of ignorance need not endure for
all after incarnation. We require only that the beings to
whom we owe justice are such that they may benefit
from the rational, self-interested choices made in the
original position. If the veil of ignorance is employed
to ensure that choices concerning a future account of
justice do not discriminate against particular social,
racial, or economic groups who stand to benefit from
just treatment, then it may also obscure the possibility
of incarnation as a mental defective or animal. The one
restriction on the scope of this incarnation would be
that only those who may benefit from just treatment
are considered (it would be senseless to speak of treating
arock or a bug unjustly).ll
Quite apart from Rawls' specific proposal, we can
approach the issue from the general perspective of the
Between the Species
nature of contracts and their results. The social contract
theory argues against including non-rational creatures
within the scope of the moral community since the
benefits of the contract are designed to accrue to only
those who agree to it. This is because the contract is
motivated out of self-interest, and all of the rules that
result prescribe behavior which serves the collective
long-term interest of the contractors.
The moral rules that result from a hypothetical
contract, while motivated out of self-interest, must
include elements such as justice and fairness whose
application does not in all cases maximize self-interest
for all the contractors. In other words, applying such
principles may require that persons occasionally forgo
their own interests in recognition of the interests of
others. Indeed, Gauthier sees that the great challenge
of formulating a rationally grounded morality lies in
defending the seemingly paradoxical claim that pursuit
of a long-term strategy of self-interest often requires
that one strikes a bargain that is less advantageous than
some alternative. Contractarian morality has excluded
nonrational creatures from the moral community
because it limits the benefits of the contract to only those
who are at least capable of participating in its formu­
lation. This assumption is captured by the passage from
Gauthier cited earlier in which he notes the impossibility
of rational cooperation between horses and humans.
So, if contractarian morality allows, even demands
each contractor to forgo maximum benefit to one's self,
all that remains is the question of whose interests the
theory allows us to consider. As a question about who
may benefit from a contract,rather than who can
participate in one, the traditional argument that non­
rational creatures are not, indeed, cannot, be covered
by the contract seems odd.
For instance, spectator sports often seem at least
partially designed to benefit non-players-those not
even capable ofplaying!-as well as players. So, too,
it is not unreasonable that the moral rules resulting from
the social contract may be designed to benefit those
who lack moral agency and who cannot, following the
analogy, play the game.
The goal of any theoretical account of moral
obligation is to determine not merely the moral status
of this or that action but of whole classes of actions;
for this reason, a satisfactory moral theory must be stated
at a relatively high level of abstraction. For contrac­
tarianism this has meant a focus not on any existing
contract but, rather, on a hypothetical contract that
24
Winter 1993
Animals, Mental Defectives, and the Social Contract
rational persons would agree to under an imaginary set
of circumstances. After all, human beings almost
certainly never explicitly agreed to any contract
outlining the set of all of our moral rights and duties.
This requirement, demanding that we look to
hypothetical, not actual, contracts, provides a theoretical
explanation for how the welfare of the nonrational can
be provided for under a contractarian framework. All
we need ask is what would a nonrational being bargain
for ifit were temporarily able? The fact that the ability
to bargain rules out any actual nonrational creature is
beside the point, since the nonrational creature imagined
is rational only while making the agreement. This is
not as farfetched as it may seem, since the temporarily
rational regularly make agreements similar to this-as,
for example, when one leaves instructions in a living
will directing how one wants to be treated if through
disease or accident one becomes nonrational. 12
Similar to this, we can imagine a (obviously
hypothetical) temporarily rational horse bargaining for
the sorts of advantages that would benefit ordinary
horses. Since ordinary horses have relatively limited
capacities to appreciate any supposed advantage, the
moral rules that result would be necessarily one-sided.
Humans would have real duties to horses, but their
duties to other humans would be considerably more
stringent and comprehensive.
There is yet another explanation for how the scope
of a contract can be supposed to include those other
than the initial contractors. It is not surprising that
rational, self-interested beings often enter into contracts
intending to benefit others as well as themselves. The
reasons for this are psychological and varied, but
perhaps the most obvious is that persons typically prefer
engaging in activities that allow for a sense of a shared
experience. We do not need to assume parental altruism
to recognize that we often engage in social behavior
precisely because, all else being equal (like the man
who refuses to drink alone or the artists and athletes
who prefer to exercise their gifts for the benefit of an
appreciative audience), we prefer sharing our well-being
with others. If we make this psychological assumption,
we see that rational agents may opt for the sort of
contract which provides the greatest benefit to all
those capable of benefiting from it. Extending the
benefits of one's good fortune to others, so long as one
does not as a consequence seriously undercut one's own
reward, is, at a minimum, quite consistent with rational
decision-making.
Winter 1993
We require further study to determine with any
precision how to balance, according to contractarian
morality, our own interests against the welfare of
the nonrational. In the case of animals, it may tum
out that depriving ourselves of the taste of their flesh
is a relatively minor sacrifice in contrast to the
benefits ceded to them. On the other hand, and more
likely, even on the contractarian account, we are
obliged to treat animals much better than we already
do. With respect to seriously defective humans, my
sense is that our treatment of them is already quite
decent. The above analysis does not recommend a
difference in treatment of them; instead, it offers an
alternative theoretical explanation that replaces mere
species-preference.
So long as we recognize that animals and seriously
defective humans are similar to us in the morally
relevant sense that they, like us, have an interest in
avoiding pain and engaging in pleasurable activities in
a stimulating environment, we can, according to the
analysis above, explain moral duties toward them.
By way ofa conclusion, in the first instance contracts
may issue from Rawlsian hypothetical circumstances
which include the possibility that the contractors will
be incarnated as nonrational beings. If so, then the rules
that are agreed on ought to speak to the interests of
those who are capable of benefiting from decent
treatment as well as those who are capable of initiating
it. Second, any plausible contractarian account of
morality conceives of the contract as hypothetical, not
actual, and there is no obvious reason why the
hypothetical circumstance conceived may not be framed
in terms that a creature would demand for itself if it
were temporarily capable of rational activity. Third,
human beings typically prefer agreements that allow
for a wider, rather than a narrower, range of benefit out
of a natural desire to share the benefits of one's own
good fortune so long as this sharing does not entail a
serious diminution of one's own welfare.
If this analysis is correct, then we ought no longer
test the plausibility of a contractarian theory ofmorality
against implications it has for seriously defective
humans. Also, the price we pay for aligning contrac­
tarian theory with intuitions concerning our duties to
defective humans forces us to accept similar duties
toward some animals. This latter conclusion, although
not strongly supported by intuition, is an inevitable
consequence of other intuitions that most of us are
reluctant to abandon.
25
Between the Species
Animals, Mental Defectives, and the Social Contract
in the just society, it would seem that they would
have to be ignorant of their species membership as
well-subject only to the qualification that they
shall have interests as participants in the just society.
Notes
1 John Rawls, A Theory of Justice (Harvard University
Press, 1971); and, David Gauthier, Morals By Agreement
(Oxford University Press, 1986).
2
12 An example from fiction may be helpful here. In
"Flowers for Algernon," a severely retarded man is
transformed into a genius as a result of taking a miracle drug,
only to gradually descend back into imbecility as the effects
of the drug wear off. Imagine the sort of demands such a
person might make while he was temporarily rational
regarding how he wanted to be treated in the future.
Rawls, p. 512.
3 The U. S. Constitution recognizes this. Allowing "cruel
and unusual punishment" is simply inconsistent with a
civilized rendering of criminal justice.
4 Rawls adds to the confusion by identifying both the
"natural duty not to be cruel" and the "duty of justice" as
natural duties, thereby reinforcing the notion that both are
basic moral notions treated by his theory.
For an extended criticism of Rawls' theory as it applies
to animals see, Tom Regan, The Case for Animal Rights
(University of California Press, 1983), pp. 163-74.
5 Gauthier,
p. 17.
6 Peter Singer, "Animals and the Value ofLife," in MOlters
of Life and Death, edited by Tom Regan (Random House,
1986), p. 366.
7 For instance, the doctrine of dominion roughly states
that God gave man dominion over all the creatures of the
earth. Humanity's superior status is justified by the possession
of an eternal soul (in the image of God), which animals lack.
As Singer points out (op. cit., pp. 346-48), even if we accept
that humans have an immortal soul that animals lack, it is
unclear why this argues for rather than against the
permissibility of taking animal life.
8 See, Jean Beer Blumenfeld, "Abortion and the Human
Brain," Philosophical Studies, Vol. 32 (1977): pp. 251-68;
Carl Cohen, ''The Case for the Use ofAnimals in Biomedical
Research," The New England Journal of Medicine, Vol. 315
(1986): pp. 865-70; and Robert Nozik, "About Mammals and
People," New York Times Book Review, (November 27,1983).
9 Michael Tooley, Abortion and Infanticide (Oxford
University Press, 1983), p. 71.
10 It is necessary to explain this in light of the fact that
undoubtedly a vast majority of people would deny it. We tend
to think that species membership is morally relevant simply
because human beings are the only moral agents of which we
are aware. This is, however, merely a contingent fact about
our experience.
11 Donald VanDeVeer ["Of Beasts, Persons, and the
Original Position," Monist, Vol. 63, no. 3 (1979): pp. 372-73]
makes a similar point. He writes:
If, then, the original position were fully neutral, its
participants would not only have to be ignorant of
the race, sex, or social position of its participants
Between the Species
26
Winter 1993